Most deaths from solid tumors are due to metastases?cancer cells breaking off the primary tumor, traveling to distant parts of the body and establishing new growths there. Patients with cancer shed millions of circulating tumor cells (CTCs) into their bloodstream every day. There is a good deal known about how malignant cells break away from the primary tumor and move through connective tissue into lymphatics and blood vessels. Almost all CTCs die in the vasculature or are eliminated by the body?s inflammatory response. However, those tumor cells that manage to extravasate out of the bloodstream and into tissues have the potential to grow as metastatic colonies that spread the cancer far beyond the primary site. Unfortunately, there is almost nothing known about this step. We have evidence that a substantial fraction of melanoma cells can cross blood vessels using the same molecular machinery that white blood cells do. Leukocytes migrate across endothelial cells (EC) without increasing vascular permeability or inducing tissue damage. They do this through a series of molecular interactions with molecules on the endothelial cells that recruit membrane from a perijunctional organelle called the lateral border recycling compartment (LBRC). Membrane from the LBRC increases the surface area of the junction allowing leukocytes to pass across without harming the EC; the junction barriers remain tight and there is no exposure of the basement membrane. Our preliminary data show that a major fraction (1/3 to ) of melanoma cells can migrate across endothelial cells. Moreover, we can selectively block their transmigration in vitro and in vivo using a novel cell-permeable inhibitor of LBRC movement. Strikingly, the number of metastatic colonies that develop in mice treated with the inhibitor is disproportionately lower than the degree to which extravasation is blocked. This suggests that those melanoma cells that transmigrate using the LBRC have a survival advantage. We hypothesize that blocking recruitment of the LBRC by melanoma cells would be an efficient way to block successful metastases. We will test our hypothesis in mechanistic studies that determine the extent to which different driver mutations affect the ability of melanoma cell lines and malignant melanoma cells derived from patients recruit the LBRC to transmigrate (Aim I). We will use our novel selective inhibitor of LBRC movement to block metastases of murine melanomas in syngeneic wild-type mice as well metastases from human patient-derived melanoma xenografts (PDX) in NSG mice (Aim II). Melanomas do not express the molecules used by leukocytes to recruit the LBRC; Aim III will determine the mechanisms used by tumor cells to recruit the LBRC. This will reveal therapeutic targets for selectively blocking metastasis without inhibiting the immune response. Preliminary data show that inhibitors of reactive oxygen species as well as endothelial EP4 receptors block melanoma transmigration in vitro, suggesting that ROS and PGE2 secreted by melanomas recruits the LBRC. We will test whether metastases from murine melanoma and PDX are reduced in inducible endothelial cell-specific TRPM2 or EP4 knockout mice.